The short answer Pick the partner whose senior people will actually do your work, who can tell you what they would measure before they tell you what they would make, and who is willing to say no to part of your brief. Everything below is detail on those three things.
Almost every "how to choose an agency" article is published by an agency, which is why they all conclude that you should hire an agency. This one does not. Three of the situations described further down are cases where you should not hire anyone at all, and we lose pitches over them regularly.
What kind of partner do you actually need? Four different things get called "an agency" in India, and picking the wrong category is the most expensive mistake in this process.
A full-service or creative marketing agency. Strategy, content, social, paid and production under one roof. Right when the pieces need to talk to each other, and when you do not have someone in-house to coordinate specialists. Wrong when you only have one clearly defined problem.
A specialist agency. Performance only, or influencer only, or production only. Right when you know exactly what is broken and it is one thing. Wrong when the actual problem is that nobody owns the whole picture, which is more often the case than founders expect.
Freelancers. A designer, an editor, a media buyer. Cheapest per hour, and genuinely the correct answer more often than agencies admit. Wrong when you do not have the time or expertise to brief and coordinate them, because that coordination is a real job and it will land on you.
An in-house hire. Right when the work is continuous, deeply product-specific, and you can attract someone good. Wrong when you need five skills and can afford one salary, which is the trap most growing brands fall into: they hire one generalist marketer and quietly expect strategy, design, copy, paid and reporting from them.
A useful test: write down the last three things that did not get done. If they span more than two disciplines, a single hire will not fix it.
The five questions that actually predict the outcome Ask these on the first call. The answers tell you more than any deck.
1. Who specifically will do the work, and will they be on this call? The oldest problem in this industry is that senior people pitch and junior people deliver. Ask for names and roles, ask who writes, who edits, who buys the media, and ask whether those people are on the call. If the answer is vague, or the room is all business development, you have learned what month three looks like.
2. What happens in month four? Most agency relationships are good for three months. The honeymoon covers the first campaign. Ask what the cadence looks like after that: who reviews performance, how often the strategy gets revisited, what happens when something stops working. A partner who has thought about month four has kept clients past it.
3. What would you measure, and what would make you say this is not working? This is the single most revealing question. Good partners answer with a small number of specific business metrics and a threshold. Weak ones answer with reach, impressions and engagement, or they cannot name a number that would count as failure. If nobody will define failure at the start, nobody will admit it later.
4. What would you tell me I do not want to hear? You are testing whether they will disagree with you before they are paid, because they certainly will not after. Any partner who cannot find a single thing to push back on in your brief is either not paying attention or is planning to sell you what you asked for rather than what you need.
5. Show me something that did not work, and what you did next. Every portfolio is a highlight reel. The useful signal is how they talk about a campaign that underperformed. You are looking for a clear diagnosis and a specific change, not a story where the client or the algorithm was at fault.
Red flags worth leaving over - **Guaranteed results.** Nobody can guarantee rankings, reach or revenue. A guarantee means either a meaningless metric or a risky method. - **Reporting that leads with reach.** Impressions are the easiest number to move and the least connected to your bank balance. If reach is the headline, ask what sits under it. - **No attribution plan before the campaign.** If nobody proposes unique links, codes or a measurement setup before the work ships, the results will be a story told afterwards, not a measurement. - **The pitch team disappears after signing.** Ask directly whether the people in the room stay on the account. - **A deck of trends with no point of view.** Anyone can list what is happening on the platforms. You are paying for a judgement about what your brand should do about it. - **Pressure to sign this week.** Discounts that expire are a sales tactic, not a business reason.
What actually moves the price Anyone who gives you a firm number before understanding scope is guessing, so treat published rate cards with suspicion. These are the factors that genuinely move a quote, and understanding them lets you interrogate any proposal you receive.
- Volume and cadence. Four pieces of content a month and twenty a month are different businesses, not the same one scaled.
- Production intensity. Founder-shot phone content, a studio shoot, and a location film with a crew and talent sit at completely different levels. This is usually the biggest single driver.
- Whether paid media is included, and whether the media budget is separate. Always check whether the fee includes the ad spend or sits on top of it.
- Creator fees. In influencer work these are typically a pass-through cost on top of the management fee. Ask which is which.
- Usage rights and exclusivity. Buying the right to run creator content as an ad, or to keep using it for a year, costs meaningfully more than a single organic post. This is where a lot of unexpected cost lands.
- Seniority and team shape. A senior-only team costs more per hour and usually less per outcome, because less work gets redone.
- Retainer versus project. Retainers buy continuity and typically a better rate; projects suit a defined launch.
The comparison that matters is not price against price. It is price against the specific scope, because two quotes for "social media management" can describe entirely different amounts of work. Ask each shortlisted partner to state deliverables per month, who produces them, and what is explicitly excluded.
When you should not hire an agency at all Three cases, and we turn work away for all three.
Your product or positioning is not settled. Marketing amplifies whatever is there. If you cannot yet say who the product is for and why they would switch, an agency will produce beautiful content for a message that is still moving, and you will pay to relearn what a few customer conversations would have told you.
Your budget only covers one month of anything meaningful. Content and paid both need iteration to work. A single month buys you one guess. Either save until you can commit to a run of three to six months, or spend the money on something with a faster payback.
You have one clear, contained problem. If you genuinely just need a better logo, a single video, or someone to fix your Meta pixel, hire a specialist freelancer. A full-service agency is overhead you do not need for a defined task.
There is also a fourth, softer case: if you have a strong in-house marketer who is only stuck on capacity rather than capability, freelancers under their direction usually beat an agency, and cost less.
How to run the selection without wasting a month 1. **Write a one-page brief before you talk to anyone.** The business problem, what you have tried, the rough budget range, and how you will judge success. Send the same page to everyone. 2. **Shortlist three.** More than three and you cannot give any of them a real conversation. Fewer and you have no comparison. 3. **Pay for a paid discovery or a small first project if you are unsure.** A short paid engagement tells you more about working together than any pitch, and it is far cheaper than a twelve-month mistake. 4. **Ask for a reference from a client who left.** Anyone can supply a happy current client. How a partner behaves at the end of a relationship is the real character test. 5. **Read the contract for the boring parts.** Notice period, ownership of assets and accounts, what happens to your ad account and creator contracts if you part ways. Make sure you own your accounts, not the agency.
A note on where we sit We are a creative marketing and social media agency in New Delhi, so we are obviously not a neutral party. What we can tell you is that the criteria above are the ones we would want a client to apply to us: senior people doing the work, measurement agreed before anything ships, and a willingness to say which parts of a brief we would not take.
If you want a second opinion on a proposal you have received from someone else, send it over. We will tell you what we would ask them, and we do that whether or not it ends with you working with us.